Breast cancer survivors confront a number of post-treatment problems including fatigue, decreased physical function, fears of recurrence, and treatment-related sequelae. Persistent fatigue, the most common and distressing problem, appears to be related in part to overactivation of the inflammatory network. Higher levels of depressive symptoms can effectively prime the inflammatory response, promoting larger NF-kappa B and proinflammatory responses to stressors. Physical activity interventions that simultaneously enhance physical performance and reduce depressive symptoms can reduce fatigue-related symptoms, as well as inflammation. This project provides the opportunity to examine mechanistic connections among fatigue, stressors, depressive symptoms, and proinflammatory cytokines both cross-section ally and longitudinally in a randomized clinical trial, building on preliminary studies that have demonstrated anti-inflammatory changes associated with yoga. A total of 200 stage I, II, and Ilia breast cancer survivors will be recruited, ages 40 and older; the women will have completed cancer treatment within the past year (except for tarnoxifen/aromatase inhibitors), and will be at least two months post surgery or adjuvant therapy or radiation, whichever occurred last. All of the women will have received chemotherapy as part of their treatment. Using a block-randomized design, participants will be assigned to either a 1 2-week (twice-weekly) hatha yoga intervention or a delayed intervention arm that receives the yoga intervention after a 6-month observation period. Differences between the two arms will be assessed at 3 months (at the conclusion of the 12-week intervention), and at 6 months (3 months after the formal intervention), controlling for any baseline (pre-randomization) measures. NF-kappa B and proinflammatory cytokine responses to a laboratory stressor will be assessed at baseline and 3 months. AIMS: (1) The primary aim is to determine if the yoga intervention will decrease inflammation, fatigue, and depressive symptoms relative to the wait-list controls. Additional aims are (2) to ascertain the extent to which the intervention modulates psychological, behavioral, and physical functioning; (3) to evaluate the relationship between depressive symptoms and the magnitude of the physiological responses elicited by a laboratory stressor, as well as the relationship of both to fatigue; and (4) to assess the extent to which the yoga intervention wiII decrease physiological stress responses.